Investigation of alternative carbon sources for the biological treatment of synthetic sulphate-laden water and mine impacted water in a linear flow channel reactor

Tawodzera, Nyasha

Investigation of alternative carbon sources for the biological treatment of synthetic sulphate-laden water and mine impacted water in a linear flow channel reactor

Thesis / Dissertation

2025

Publisher

University of Cape Town

Department

Department of Chemical Engineering

Faculty

Faculty of Engineering and the Built Environment

Abstract

South Africa grapples with the generation of acid mine drainage (AMD), which adversely affects surface and ground water quality. Existing treatment methods typically treat the acid and heavy metal components of AMD but often fail to meet sulphate removal standards, necessitating additional polishing steps that add to expense. They also carry drawbacks such as high operational costs and metal sludge generation for active treatment and reduced process control and the need for large land areas for installation for passive treatment. These treatment methods are not cost-efficient when treating low-volume, circum-neutral wastewater. Biological sulphate reduction (BSR) offers a sustainable alternative for sulphate removal and is applicable to partially treated AMD as well as circum- neutral, mine-impacted water. Coupled with partial sulphide oxidation, it also has the potential to convert sulphate to elemental sulphur, touching on waste valorisation as sulphur is a value-added product. However, BSR systems require supplementation with organic carbon and are characterised as slow, while partial oxidation of sulphide is difficult to control in many reactor systems; these are key drawbacks in the economic feasibility of these processes. A semi-passive linear flow channel reactor (LFCR) which simultaneously reduces sulphate to sulphide using sulphate reducing bacteria (SRB) and partially oxidises the sulphide formed to elemental sulphur using sulphur oxidising bacteria (SOB) within a floating sulphur biofilm, was developed at the Centre for Bioprocess Engineering Research at the University of Cape Town, South Africa. It may be operated as a one- or two-stage reactor system, with the second stage providing additional surface area for partial sulphide oxidation. During its development, carbon fibres were added to enable biomass retention and thereby enhance reaction rates. For this study, the primary reactor was further modified to include baffles for improved directional flow and enhanced contacting and the carbon microfibre biomass support was replaced with polyurethane foam (PUF) previously shown to effect efficient biomass retention. A secondary reactor was included to provide 30% of operating volumes with no baffles; it was connected in series to increase surface area for sulphur recovery. The baffled hybrid linear flow channel reactor (BaH-LFCR) was supplied with an organic substrate, lactate, and was tested for its treatment of a synthetic sulphate laden feed. Lactate has been shown to be highly effective carbon source and electron donor for sulphate reduction but is expensive and not available at sufficient scale or low enough cost to be a cost-effective option at an industrial scale. The synthetic feed, unlike AMD from the field, was nutrient-rich with a stable, neutral pH that promoted SRB function. This study investigated the selection and use of an alternative, cheap and readily available carbon source and electron donor for BSR as well as the treatment of AMD from the field in the BaH-LFCR. Alternative carbon sources investigated included molasses, acetate, honey and algal biomass; each have a high chemical oxygen demand making them potentially suitable organic carbon sources for BSR. Honey and algal biomass can be produced on-site enhancing availability and negating transportation costs. As a byproduct of the sugar industry, the equivalent COD as molasses costs less than 0.1% of that of lactate. Acetate is a byproduct of most fermentation processes making it readily available. Field AMD presents several challenges for BSR due to its acidic nature, lack of nutrients, and potential toxins. To address these challenges, the AMD was characterised and pre-treated to increase the pH before introduction into the BSR system. Use of an alternative substrate and real AMD from the field demonstrates the ability of the LFCR to achieve real-world application and implementation. Three small-scale reactor configurations were tested with lactate and sulphate-laden feed: a 1 L fed-batch Schott bottle, a 93 mL continuous mini column, and a 1 L continuous Schott bottle. Continuous reactors performed poorly due to oxygen ingress, achieving only 43.4% sulphate conversion in the Schott bottle and no conversion after 23 days in the mini column. The fed-batch reactor demonstrated better stability with 78.1% conversion. Oxygen ingress impact was found to be inversely proportional to reactor size. Based on superior stability and conversion efficiency, the fed-batch reactor was selected for carbon source testing. The four alternative substrates were tested against lactate, as base case, in the fed-batch reactor. Molasses showed the highest performance among the alternative carbon sources, achieving 82.7% sulphate conversion and producing the highest sulphide concentration. In contrast, honey and algal lysate performed poorly, with average sulphate conversions of 7.4% and 14.1% respectively. The poor performance of honey was attributed to its antimicrobial properties and acidic nature. Poor performance of algal lysate likely resulted from its composition of predominantly unusable COD which would require fermentation to produce more accessible compounds. Acetate had an average conversion of 41.4% which was approximately half the conversion achieved using molasses as a carbon source. VFA analysis revealed that molasses was the only substrate where all measurable sugars and VFAs were consumed by the end of each cycle, as indicated by HPLC analysis. Additionally, molasses contained fermenting microorganisms that converted sugars into small concentrations of lactate, enhancing sulphate reduction. These fermenters were introduced into the BSR system along with the molasses substrate. Before introducing AMD from the field to the BaH-LFCR, various AMD pretreatment methods were evaluated using lactate-fed batch reactors. Four AMD conditions were tested: untreated AMD, lime-treated AMD, lime and sulphate-treated AMD, and lime-treated sterilised AMD. The untreated AMD batch showed the lowest sulphate conversion (9%) due to its acidic pH, which inhibited SRB activity. The highest conversion of 89% was achieved with lime-treated, sterilised AMD. Sterilisation eliminated competition for the carbon source between native microorganisms present in AMD and the SRB, resulting in enhanced sulphate conversion. Three experiments were conducted in the BaH-LFCR system to evaluate its performance using different combinations of carbon sources (lactate vs. molasses) and feed solutions (synthetic Postgate media vs. pre-treated AMD from the field). The first experiment established a base case using lactate and sulphate-rich synthetic feed to determine the optimal hydraulic residence time (HRT). At the optimal 3-day HRT, this base case achieved 64.8% sulphate conversion and the highest volumetric sulphate reduction rate (VSRR) of 0.187 mmol/L.h in the primary reactor, nearly two-fold that achieved previously in the LFCR. The second experiment, using lactate with partially treated AMD, achieved the highest sulphate conversion of 87.4% and the second-highest VSRR (0.216 mmol/L.h) in the primary reactor. It had 41.6% of the sulphur entering the system through the feed converted to sulphur. Sulphur formation was observed to decline, likely due to the development of a thin, impervious surface film hypothesised to consist of calcium crystal complexes. This film may have impaired oxygen di`usion at the air-liquid interface more severely than the typical floating sulphur biofilm (FSB), thereby reducing the efficiency of sulphide oxidation to elemental sulphur. The third experiment, combining molasses with partially treated AMD, achieved second highest sulphate conversion of 85.6% in the primary reactor. Overall sulphate conversion however dropped to 27.2% due to extensive re-oxidation in the secondary reactor. This re-oxidation linked to poor FSB formation and limited carbon availability in the secondary reactor. However, in the primary reactor the molasses configuration achieved the highest proportion of expected sulphide converted to sulphur of 30.7%, with a comparable expected sulphide amount of 518 mmol. The synthetic feed + lactate experiment achieved approximately only 9.4% sulphide conversion in the primary reactor, with an expected sulphide amount of 428 mmol while the AMD + lactate experiment had the highest expected sulphide amount of 543 mmol. In summary, the introduction of partially treated AMD into the LFCR system on a lactate carbon source not only maintained but enhanced system performance, achieving the highest sulphate conversion despite lacking the additional nutrients present in the SRB-specific feed. While using molasses as a complex waste stream carbon source achieved high sulphate conversion in the primary reactor, sulphur recovery was compromised due to re-oxidation at the primary reactor effluent port and because of limited carbon availability there was poor sulphur formation and high sulphate concentrations. The effective treatment of circum-neutral, sulphate-laden mine-impacted water using a readily available, cost-effective substrate demonstrates the system's suitability for industrial-scale deployment.